INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

Plasmodium vivax, one of the two major human malaria parasites, accounts for a large proportion of the world's clinical malaria infections due to its prevalence in much of Asia and South America; in Sri Lanka, P. vivax is responsible for 60 to 80% of the 300,000 to 400,000 clinical malaria infections which occur annually (1). Despite its wide reputation as the benign malaria parasite, it is not uncommon for it to lead to a severe and complicated pathology, including cerebral malaria and death (19, 25, 43). Thus, although it has been the subject of much less study than Plasmodium falciparum, due mainly to the lack of a continuous in vitro culture system, the threat to health from P. vivax in parts of the world where malaria is common is high. In addition, there is growing resistance of P. vivax parasite strains to chloroquine in many countries (12, 16, 17, 38). The situation therefore calls for extending to P. vivax the vaccine development efforts which are currently focused almost exclusively on P. falciparum malaria.

Our own efforts at vaccine development for P. vivax have focused on the major merozoite surface protein MSP1 (9, 29, 42), which is a leading candidate for an asexual blood stage vaccine against P. falciparum malaria (10). MSP1 is synthesized in hepatic and erythrocytic schizonts as a 200-kDa membrane-bound precursor (reviewed in references 6 and 21), which, around the time of merozoite release, is processed in P. falciparum in two steps leading first to 42-kDa and then to 19-kDa glycosylphosphatidylinositol-anchored C-terminal molecules. An immunization trial with the native MSP1 protein of P. falciparum conferred a highly protective immunity against homologous challenge in Aotus monkeys (39). Subsequently, protection was also elicited by Escherichia coli-expressed recombinant C-terminal MSP1 p19 analogs in rodent malaria models (7, 27). MSP1 p19 C-terminal recombinant antigens of P. falciparum produced in yeast (26) and p42 antigens produced in baculovirus (3) have also given protection in monkeys, indicating their potential as components of a subunit vaccine.

One of the greatest impediments to malaria vaccine development concerns the lack of in vitro correlates and suitable in vivo models of protective immunity against malaria in humans. Thus, the efficacy of candidate vaccines is assessed only at advanced stages of clinical testing, and the process from selection of the antigen through development of the product could take up to several years. This uncertainty of efficacy, combined with the high costs involved in product development and testing, has deterred potential investors from the field and slowed considerably the process of malaria vaccine development.

We describe here the use of a natural host-parasite system which appears to be highly analogous to human P. vivax malaria. The toque monkey, Macaca sinica, is one of the natural hosts of P. cynomolgi; it lives in regions of Sri Lanka which include the enzootic foci where the parasite abounds (11, 35). The analogy between the simian parasite P. cynomolgi and its human counterpart P. vivax is substantiated by biological (5, 8), genetic (13-15, 30, 44), and evolutionary (18) evidence (reviewed in references 28 and 31). Studies that we have carried out with this natural nonhuman primate system involving transmission-blocking immunity (34), relapses (unpublished data), and pathogenesis (33) have found direct application to human P. vivax malaria (31). P. cynomolgi and P. vivax MSP1 C-terminal recombinant proteins produced in the baculovirus expression system are highly homologous and appear to reproduce the physical and antigenic characteristics of the native proteins (22, 28). This report presents a highly successful vaccine trial with the toque monkey, showing excellent immunogenicity and protective efficacy of the P. cynomolgi MSP1 recombinant proteins.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Animals. Adult toque monkeys (M. sinica sinica) were captured from their natural habitats in the dry zone of Sri Lanka and were quarantined in the animal house for 1 month prior to experimental use. All animals were screened for the presence of a current malaria infection by microscopic examination of thick blood films and for previous exposure to malaria by detection of serum antimalarial antibodies by the indirect immunofluorescence assay (IFA). An animal was considered naive with respect to malaria when thick blood films failed to detect parasites for 7 consecutive days and anti-P. cynomolgi antibodies were undetectable. We have shown that such animals are free of subpatent malaria infections because they are incapable of transmitting malaria infections to other naive recipients by blood passage (20, 32).

Recombinant immunogens. P. cynomolgi MSP1 recombinant proteins corresponding approximately to the P. falciparum MSP1 42- and 19-kDa C-terminal processing fragments (28, 29) were produced in the baculovirus expression system and purified by immunoaffinity chromatography with monoclonal antibodies as described previously (22).

Immunization of animals. Fifteen malaria-naive animals comprising five groups of three animals each, matched by sex and weight (Table 1), were used for the trial. Animals in the first three groups (groups I, II, and III) were vaccinated with either p42, p19, or a combination of the two antigens, respectively, and those in group IV were vaccinated with normal saline; group V served as unvaccinated controls. Each inoculation consisted of 100 µg of the p42 antigen and/or 35 to 50 µg of the p19 antigen administered subcutaneously in three doses at 4-week intervals. Each dose was suspended in 0.5 ml of normal saline and emulsified in an equal volume of adjuvant consisting of Freund's complete adjuvant (FCA) and Freund's incomplete adjuvant (FIA) in proportions of 1:1 (first immunization) and 1:4 (second immunization). Only FIA was used for the third and last immunization. Immunized control animals in group IV were administered normal saline with the same adjuvant protocol. No clinically adverse reactions to vaccination were evident in any of the animals, even at the sites of subcutaneous vaccine injection. Serum samples were taken prior to the first immunization and 2 to 3 weeks after each immunization.

Challenge infections and evaluation of protective efficacy. Animals in all five groups were challenged 4 weeks after the third immunization with an inoculum of 2 × 10⁵ asexual blood stage P. cynomolgi ceylonensis parasites obtained from a donor monkey. Parasites are maintained in the laboratory by blood and mosquito passage and cryopreservation (32, 34). Commencing from the fourth day after challenge, all 15 animals were screened for asexual blood stage parasites by microscopic examination of Giemsa-stained thick and thin smears prepared from an earprick. Blood film examinations were conducted daily for 6 weeks and weekly up to 8 weeks postchallenge.

ELISA. Ninety-six-well microtiter plates (Nunc, Roskilde, Denmark) were coated with 0.1 µg of either p19 or p42 antigen per ml and incubated at 4°C overnight, and enzyme-linked immunosorbent assays (ELISAs) were carried out as previously described (37) with horseradish peroxidase-conjugated goat anti-human immunoglobulin (1/1,000 dilution) (DAKO Immunoglobulins, Copenhagen, Denmark). The optical density was read at 492 nm and the ELISA end point was taken as the highest serum dilution which gave an optical density greater than twice that of control sera.

IFA. P. cynomolgi schizont-infected erythrocytes were purified from infected blood by a Percoll density gradient method as described previously for P. vivax (23). The infected-erythrocyte suspension was spotted onto 12-well antigen slides, air dried, and stored at 70°C. The IFA was performed as previously described (42) with fluorescein isothiocyanate-conjugated goat anti-human immunoglobulin G (Organon Teknika Corp., West Chester, Pa.).

SDS-polyacrylamide gel electrophoresis and Western blotting. Purified P. cynomolgi schizonts were extracted in sodium dodecyl sulfate (SDS) sample buffer with reducing agents, the polypeptides were electrophoretically separated on SDS-10% polyacrylamide gels and transferred to nitrocellulose paper by electroelution, and the Western blot was processed as described previously (36).

Statistical analysis. Log transformations of parasitemias were obtained, and comparison of means was done by analysis of variance.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Immunogenicity. Both the recombinant p19 and p42 antigens, either alone or in combination, were highly immunogenic, since animals in all three MSP1-vaccinated groups developed very high antibody titers against the vaccinating antigens as detected by ELISA (10⁶ and above) and against P. cynomolgi schizont-infected erythrocytes as measured by IFA (>10⁵) (Table 1). Although antibody titers were already quite high after a single injection of the recombinant antigen, these titers increased in postimmunization sera in most animals by about 10-fold or more from the first to the last injections (Fig. 1). IFA titers of these sera also showed a similar increase through consecutive immunizations (data not shown).

View larger version (20K): [in this window] [in a new window]   FIG. 1.   ELISA end point titers of all pre- and post-immunization sera against p42 (A) and p19 (B) recombinant proteins.

Protection against parasite challenge. Following challenge, adjuvant-treated control animals of group IV developed P. cynomolgi parasitemias similar to those in the unvaccinated controls. All animals in both groups had patent infections for 39 to 54 days, although the parasite densities in the adjuvant-treated controls were significantly lower than those in the unvaccinated controls (P = 0.0006). These adjuvant-treated controls also experienced lower peak parasitemias (0.1 to 0.6%) than unvaccinated controls (0.4 to 3.8%) (Fig. 2A; Table 1), indicating some effect of adjuvant alone. In contrast, the three animals immunized with p19 (group II) developed either no parasitemia at all (T429) or a transient infection of 0.002% which was patent for either 1 (T427) or 2 (T426) days (Fig. 2C; Table 1). All but one of the six animals in the other two vaccinated groups were similarly protected, with peak parasitemias of 0.002 to 0.02% and 0- to 2-day patencies (Fig. 2B and D; Table 1). One animal in the p42-vaccinated group (group I, animal T435) developed a parasitemia that peaked at 0.06% and persisted for 17 days. The parasite densities from day 5 to 32 of challenge infections in each of the three vaccinated groups were significantly lower than those in both the adjuvant-treated control group and the unvaccinated control group (F = 23.7788; P < 0.0001).

View larger version (25K): [in this window] [in a new window]   FIG. 2.   Courses of blood parasitemia following challenge infection with P. cynomolgi in naive unvaccinated (group V) and adjuvant-treated control (group IV) animals (A) and in animals immunized with p42 (group I) (B), with p19 (group II) (C), and with both p42 and p19 (group III) (D).

View larger version (20K): [in this window] [in a new window]   FIG. 3.   Courses of blood parasitemia following the first and second challenge infections with P. cynomolgi in naive control animals (A), in adjuvant-treated control animals (group IV) (B), and in p19-vaccinated animals (group II) (C).

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Vaccination with the baculovirus-expressed p42 and p19 C-terminal P. cynomolgi MSP1 recombinant proteins resulted in a very high antibody response and very effective protection against a challenge infection in eight of nine animals, with an apparently sterile immunity induced in two. The low transient blood parasitemias of challenge infections seem to have been completely cleared, since parasite DNA could not be detected by PCR 34 days after challenge. The three animals in the p19-vaccinated group were also able to resist a second challenge 6 months later.

The high degree of protection elicited in this trial could be due to the nature of the recombinant proteins themselves, which were produced in a higher-order eukaryotic system and appear to reproduce the complex conformational structure of the native protein. We have, in fact, shown that the reactivity of human antisera from P. vivax-immune individuals is completely dependent on the integrity of reduction-sensitive conformational epitopes present in the p19 recombinant protein (22).

It is noteworthy that the smaller MSP1 p19 antigen containing only the two EGF domains conferred at least as good protection as the larger p42 antigen, suggesting that this part of the molecule is sufficient to generate protective immunity. This clearly indicates that the p19 antigen must have its own functional B- and T-cell epitopes that mediate a strong effective immune response without the addition of extraneous non-MSP1 peptides. We have previously shown that the baculovirus-expressed recombinant MSP1 p42 and p19 antigens have distinct physical properties and appear to display epitope subsets that are specific to each (22). In addition, the p42 protein appears to have a hypervariable region which may play a role in immune evasion, assuming that it is the focus of an effective antiparasite immune response (28). The presence of this region in a subunit vaccine might therefore be detrimental in a malaria-endemic environment where the risk of exposure to polymorphic variants in the hypervariable region may render vaccination with a single variant less effective.

The potent protective immunity observed in this trial might have been contingent upon the exceptionally high antibody titers that were elicited in the vaccinated animals by using Freund's adjuvant. Indeed, the serum of the least protected vaccinee (T435) showed the slowest rise in ELISA titers against both the recombinant antigens, even though the titers were still quite high (10⁶). In the primate P. falciparum trials involving recombinant MSP1 antigens, the protection observed did not appear to be dependent on antibody titers (3, 26). However, if this level of vaccine-induced protection does indeed require the presence of such high antibody titers, then the challenge ahead for preparation of an analogous human P. vivax vaccine will be to replace the potent Freund's adjuvant. Indeed, parasitemias after challenge in the Freund's adjuvant-treated controls were generally somewhat lower than those in the unvaccinated controls, suggesting that this strong adjuvant may contribute nonspecifically to the immunity generated in this model. Nevertheless, preliminary results do in fact suggest that significant protection can be obtained in this host-parasite system with the p19 antigen by using alum, which is the only adjuvant currently acceptable for human use, even though antibody titers were not as high as those observed with FCA and FIA (36a). The p19-vaccinated animals were still protected against a second challenge 6 months later, despite very low or undetectable parasitemias following the first challenge infection. Similar resistance to a second homologous challenge in the adjuvant-treated controls is attributable to immunity induced by much higher sustained parasitemias observed after the first challenge. Several observations indicate that protection against homologous Plasmodium strains can be induced by a single complete blood stage infection (see below). We have evidence to suggest that the sustained protection of the p19 vaccinees to a second challenge was due specifically to an anti-MSP1 p19 immune response which may have been either sustained following the first immunization itself or, more likely, boosted by a minimal exposure to parasites during the first challenge. If the latter is the case, it implies that vaccine-induced immunity could be boosted by natural infection, a feature which could be of considerable importance for the deployment of an MSP1 p19-based vaccine in malaria-endemic areas.

Protection as complete as that elicited in this trial has rarely, if ever, been achieved in primates with a recombinant subunit malaria vaccine against blood stage parasites. These results are now being confirmed in a trial with larger numbers of animals. Consistent success has, however, recently been achieved by immunization with baculovirus-expressed recombinant products of this antigen, both in P. vivax with MSP1 p42 and p19 in a nonsplenectomized squirrel monkey (Saimiri sciureus boliviensis) model (1a) and in P. falciparum with MSP1 p42 in Aotus monkeys (3), even though both used FCA and FIA as adjuvants. A recent successful clinical trial of a preerythrocytic stage recombinant malaria vaccine in conjunction with an oil-in-water adjuvant containing monophosphoryl lipid A and QS21 (41) has considerably heightened the prospects for replacing Freund's adjuvant for human use.

The results of this trial have demonstrated that a high degree of protection against a malaria blood infection can be achieved in this natural primate host-parasite system by vaccinating with a recombinant subunit vaccine. Two points need to be considered in assessing the relevance and stringency of this model for vaccine evaluation. First, in the course of an uninterrupted challenge infection of P. cynomolgi in the toque monkey, parasitemias peak at less than 5% and last for at least 44 days, and probably more than 58 days, after which they are self-cured. These characteristics are similar to those of untreated P. vivax infections in human volunteers, which reached parasitemias of less than 10% and remained patent for 60 days or more prior to being self-cured (4).

Second, a complete P. cynomolgi blood infection in the adjuvant-treated control animals following the first challenge led to almost complete protection against a second challenge with the homologous strain 6 months later. Although this might suggest that protective immunity is too easily achieved in this system for it to be a valid immunization model for human malaria, two previous vaccine trials that were conducted with this system, using some of the same antigens under different conditions, failed to provide any protection, for a variety of possible reasons (36a). However, more important, it is likely that protection against a homologous strain of P. vivax in humans can also be induced by a single blood infection provided that the infection is allowed to run its natural course without drug treatment. Evidence for this comes from reports of early investigators who studied malaria infections which were therapeutically induced in neurosyphilitic patients (2, 24). Untreated P. vivax infection gave rise to complete clinical protection against a subsequent challenge with the identical strain but not with a different strain of the same species (2); since no parasitological observations were made in these studies, it is not possible to determine the extent to which antiparasite protection was achieved by these infections. Similar studies reported by Jeffery (24), in which all of the induced infections were drug cured, did not give rise to such a striking immunity, implying that the early termination of infection interferes with the development of immunity. The results of these early experiments also emphasize that the difficulty in acquiring natural protection against malaria must be largely attributable to polymorphism in target antigens or epitopes of protective immunity among natural parasite populations.

To test a malaria vaccine destined for use in humans, an analogous natural primate malaria may be superior to artificial experimental systems based on the human malaria parasites themselves in unnatural hosts. Nevertheless, the real relevance of the results reported here will be determined only by clinical trials with the P. vivax MSP1 recombinant analogs. In the meantime, however, this system will serve to address several important questions raised by this promising trial. These include, principally, the need for testing candidate vaccine antigens under statistically valid conditions and the search for effective adjuvants which are acceptable for human use.